Gaia Lab on Gaia Lab · Blog

Serving models and looking after the cluster: the infrastructure behind the other lines

Tue, 22 Sep 2026 14:15:00 +0200

Ninth and final instalment of Cluster X-ray. Here we group what does not fit into a research line but shows up in almost everyone’s scripts.

Serving models so that others can work #

Several lines in the series need a large model up and running: a judge to score responses, a synthetic data generator, a coding assistant. On a cluster with one GPU per node, the natural way to serve it is a Slurm job that brings up vLLM, waits for it to respond, registers with a shared proxy and stays there until its time runs out. On finishing, it deregisters: the model appears in and disappears from the catalogue along with the job.

GNNs for a lagoon's chlorophyll: does the graph add anything?

Tue, 22 Sep 2026 14:00:00 +0200

Eighth instalment of Cluster X-ray. In the overview post this line appeared as intrusion detection. The scripts tell a different story: the graph networks are environmental, and intrusion detection arrives in September as a pilot.

The question #

Chlorophyll concentration is the most direct indicator of the state of a coastal lagoon: it anticipates eutrophication episodes and blooms. It can be estimated from satellite, but the relationship between reflectance and chlorophyll depends on the place, the moment and what is happening around it. The line asks whether a graph neural network that connects points of the lagoon in space and time predicts better than tabular models, and, above all, whether the graph contributes real information or merely acts as a smoother.

A risk classifier for minors that fits in a phone

Tue, 22 Sep 2026 13:45:00 +0200

Seventh instalment of Cluster X-ray. A safety line of another kind: that of minors using a phone.

The question #

Detecting risk situations in a minor’s conversations (harassment, grooming, sexual content, isolation, threats) is possible with a language model. Doing it on a server means sending the conversation off the phone, and that is precisely what you do not want to do with a minor’s private life. The line asks whether a model small enough to run on the phone itself can classify risk with the necessary reliability, and how much is lost in compressing it until it fits.

Dental segmentation and 100 seeds: how much a model changes by sheer luck

Tue, 22 Sep 2026 13:30:00 +0200

Sixth instalment of Cluster X-ray. After the forests , the most frugal line in the archive: twenty GPU-hours in four months and a methodological question that many papers publish without answering.

The question #

Automatically segmenting structures in dental images, separating the region of interest from the background pixel by pixel, is the basis of any subsequent measurement or assisted diagnosis. The literature compares architectures with a single run per model, and the differences between them are often on the order of one point. The line asks two things: which architecture segments this type of image best, and how much of the difference between two models is real and how much is the lottery of initialisation.

Trees one by one from LiDAR point clouds: splitting the forest into individuals

Tue, 22 Sep 2026 13:15:00 +0200

Fifth instalment of Cluster X-ray. After the proteins , another line unrelated to language: forests.

The question #

A LiDAR scan of a forest plot returns millions of unlabelled points. To estimate above-ground biomass, and with it the stored carbon, you need to know which points belong to which tree: splitting the forest into individuals. It is a 3D instance segmentation problem, hard wherever the crowns touch. The line asks which architecture does it best, how much the result depends on the luck of the initialisation, and which information in the cloud really matters.

4,400 protein complexes with Boltz-2: predicting how two proteins bind

Tue, 22 Sep 2026 13:00:00 +0200

Fourth instalment of Cluster X-ray and the line with the fewest jobs: fifteen. Also one of those that reserves the most GPU, because each job runs for days.

The question #

Knowing whether two proteins interact, and how, is one of the central questions of structural biology: it explains signalling pathways, drug targets and the effect of mutations. Determining it experimentally costs months per pair. The line uses structure prediction to estimate, at the scale of thousands of pairs, the structure of the complex that two proteins form and the confidence that the binding is real.

Measuring quantized LLMs: what is lost when an open model is compressed

Tue, 22 Sep 2026 12:45:00 +0200

Third instalment of Cluster X-ray. It is the best example of what the overview post said: the cluster today is above all a measurement laboratory. Not a single parameter is trained here.

The question #

Open models are distributed quantized: the same weights compressed to 8, 6, 5, 4, 3 or 2 bits so they fit on a small card or a laptop. The community publishes hundreds of variants, but rarely with a comparable measure of what is lost. The line asks how well each quantization level of each family really performs on tasks that matter for using the model as a tool, and at what speed and energy cost it does so on each card.

A council of small models against a 120B one: automated jailbreaking and who judges it

Tue, 22 Sep 2026 12:30:00 +0200

Second instalment of Cluster X-ray. The previous instalment fine-tuned a model so it would refuse what is harmful. This line does the opposite: it automatically attacks a large model and counts how many times it gives in.

The question #

Automated jailbreak attacks usually use an attacker model that refines a prompt turn by turn. The line asks whether a group of small models that deliberate attacks better than any of them on its own, and, with equal weight, how to measure success without fooling yourself, because the same attack can look successful or not depending on who judges it.

Self-alignment: how much safety to teach a language model without making it useless

Tue, 22 Sep 2026 12:15:00 +0200

First instalment of Cluster X-ray, the series that goes line by line through what really runs on the ANTS group’s cluster . We start with the first of three lines on language model safety.

The question #

Fine-tuning a language model so that it refuses harmful requests has a cost: if it is taught too thoroughly to say no, it starts refusing harmless things too, and loses usefulness. The line studies that balance: what proportion of safety data has to be mixed with the utility data, and whether the model itself can generate its own refusal examples and regulate the dose on its own.

What really runs on our cluster, read from the scripts themselves

Tue, 22 Sep 2026 10:00:00 +0200

A cluster’s job history is usually a wall of six-digit identifiers and cryptic names —mcpfw_cc, eval_dynamic_pid, boltz2— that say everything to whoever launched them and nothing to anyone else. But if you keep the scripts, the wall turns into a document: people write down why they do things, especially when a batch costs money and a night of waiting.

Welcome to the Gaia Lab blog

Fri, 18 Sep 2026 00:00:00 +0200

With this post we open the Gaia Lab blog. Before publishing deployment notes, results and guides, it is worth putting in writing who we are, what we do and what anyone following these pages can expect.

What Gaia Lab is #

Gaia Lab is the 5G and B5G technology experimentation laboratory of the ANTS group, part of the Department of Information and Communications Engineering at the University of Murcia. Our motto, “Driving connectivity”, sums up the intent: to design, deploy and measure communication networks on real infrastructure, not just in simulation.

How to publish a post on this blog

Thu, 17 Sep 2026 10:00:00 +0200

This blog is a static site generated with Hugo and deployed on GitLab Pages. Each post is a Markdown file inside its own folder, and publishing means merging that file into the main branch. At Gaia Lab we have gathered the complete steps here so that any member of the team can write and publish without depending on anyone else.

The experimental vehicle: a 5G lab on wheels

Tue, 15 Sep 2026 10:00:00 +0200

At Gaia Lab, the 5G and B5G technology experimentation laboratory of the ANTS group at the University of Murcia, we have put into service a multidisciplinary experimental vehicle for several research lines: 5G and B5G networks, vehicular communications, perception, IoT and edge computing.

Why a lab on wheels #

Many applications that depend on connectivity can only be evaluated on the move and on real terrain: high-resolution video, digital twins at the edge, cooperative communications or georeferenced multi-sensor data. We needed a mobile edge for the Espinardo Campus’s private 5G network, equipped with sensors, radios and computing.

What it connects to: the 5G and B5G experimentation infrastructure of the Espinardo Campus

Thu, 10 Sep 2026 10:00:00 +0200

When we present the experimental vehicle we usually dwell on what it carries on board. In this post we turn the question around and explain what it connects to: the vehicle is the mobile edge of the 5G and B5G experimentation infrastructure that we at Gaia Lab, within the ANTS group at the University of Murcia, have deployed on the Espinardo Campus. We describe its five building blocks and how the vehicle closes the loop.

About

Mon, 01 Jan 0001 00:00:00 +0000

What Gaia Lab is #

Gaia Lab is the 5G and B5G technology experimentation laboratory of the ANTS group at the University of Murcia. Our motto, Driving connectivity, sums up what we do: deploying, measuring and putting next-generation networks to the test on real infrastructure, and carrying that connectivity all the way to the edge, whether that is a sensor, a roadside unit or a moving vehicle.